24/7 HR Sampling

I too wear a chest strap while working out and you'll get no argument from me that an optical HRM can't compete. Though some may be better than others, no one makes an optical HRM that'll give you readings like a chest strap while doing exercises like one would do in a cross fit workout. You'll definitely see a difference while "resting" or "recovering" between exercises, intervals, or sets. Outside of a workout, an optical HRM can have fairly accurate readings while just performing everyday activities. I believe the consensus is the FR235 does a good job at this because sampling is done based on heart rate increases and movement. The sampling we're referring to is your resting heart rate; while you're actually asleep or when you're relaxed to a point were you have little or no movement.

Does the FR235 sample hr any differently with the chest strap connected?

Interesting Tim. How did you capture those data points?

You have to either copy the fit file from your device (GARMIN/MONITOR) before syncing. Or after syncing you can export them on the HR page from Connect. Then use the fitToCSV.bat that is included in the FitSDK you can download from thisisant,com (drag and drop fit files on the bat).

Here's a little "trick" to try. Put your 235 into HR broadcasting mode (from the HR widget page - press and hold the up arrow to get to that menu). Overnight I got a load of values as below! Sample times range from 1 to 15 mins.

Interesting, do you have smart recording enabled?

Another interesting thing is that is says my current RHR is 36 yet the lowest recorded value overnight is 39 so is it taking more points for RHR but just not recording them all?

I think I noticed this once as well. But often after waking up looking at the HR widget I see the opposite, as the minimum HR being lower than what the RHR says.

It sounds like your test confirmed what Philippe's test results were. Now it has me wondering about my watch since movement was the only thing that triggered my sensor light. It could be I have a unit that's acting up, or maybe the difference between 3.10/2.20 and 3.20/2.30(40) is the earlier sampled based on movement and the latter on a time frame? There's another 3.10 user who going to test theirs tomorrow. Now I'm really anxious to see their results.

I had the impression that with 3.10 the sensor was more "sensitive" to movement than now, though it still does take more measurenements when walking around more.

I got my 235 a few days ago and was not really aware of the 'problem' (and it IS a problem) with the hr sampling rate. I did a cross fit workout today and wore my old Polar hr monitor and the 235. Readings from the 235 were a joke. Forget about accuracy during the actual workout (can't really compare a chest strap to an optical sensor), but during the rest periods there was a huge difference. At one point my heart rate on the polar was 141 and on the 235 it read 93. It took over a minute to get the two back in sync. I guess this is a great fitness tracker if your heart rate doesn't change much over the course of a day...but mine does. If I can't get an option for faster sampling I'm sending it back.

Optical sensors are no good for those type of workouts. My guess is that flexing wrist and muscles in your forearm impedes bloodflow that much, so that the sensor cant pick up a clean signal. When I say rest, I mean daily rest, like sitting or laying, not rest between reps in a workout.

Does the FR235 sample hr any differently with the chest strap connected?

This article can explain a lot better than I could. There's also a separate link under the "Accuracy" section that has some good info. Note the FR235 wasn't tested.
http://www.wareable.com/sport/heart-rate-monitors-chest-straps-v-wrist

Good morning Philippe. I tried the test again this morning. Same results. No matter what I try my 235 my will not light up unless there is some type of movement. Ian is running 3.10 and he's going to test today. It'll be nice to see if the 3.10 version acts differently, or if I'm having an issue with my device. Everything thing fine though. I just can't determine how often my is sampling at rest.

Well, it IS necessary to move it, also with newer fw. If I didn't give it a wiggle every few minutes or at least every 15-30 minutes) the light would not turn on. Thats why you have to move it from time to time. The least you should do is every 15 minutes as this is the expected sampling frquency on 3.10.

If you don't give it movements it won't turn on. And if you don't give it a reading ONCE it has turned on the test will also be different from mine.

Understood. I tried movement every 5 minutes with a reading for 30 minutes straight. After the 30 minutes I moved it every 5 -15 minutes just enough so I believe the watch thought I was wearing it, but not enough to trigger a light. Once I reach the 2nd 30 minutes of my test, the light doesn't come back on. I would think I'd get at least one sampling either 15 or 30 minutes after the last one. Maybe I'm just a poor pupil and I'm doing something wrong. :(

Understood. I tried movement every 5 minutes with a reading for 30 minutes straight. After the 30 minutes I moved it every 5 -15 minutes just enough so I believe the watch thought I was wearing it, but not enough to trigger a light. Once I reach the 2nd 30 minutes of my test, the light doesn't come back on. I would think I'd get at least one sampling either 15 or 30 minutes after the last one. Maybe I'm just a poor pupil and I'm doing something wrong. :(

Strange... maybe you shpuld try in a car someday like darren, or maybe you can wear it very loose so one can see the lights turning on. I've tried to put it on my wrist upside-down but couldn't close the lock then...

You have to either copy the fit file from your device (GARMIN/MONITOR) before syncing. Or after syncing you can export them on the HR page from Connect. Then use the fitToCSV.bat that is included in the FitSDK you can download from thisisant,com (drag and drop fit files on the bat).

Interesting, do you have smart recording enabled?

I think I noticed this once as well. But often after waking up looking at the HR widget I see the opposite, as the minimum HR being lower than what the RHR says.

I had the impression that with 3.10 the sensor was more "sensitive" to movement than now, though it still does take more measurenements when walking around more.

Yes it was taking the most recent FIT file in the MONITOR folder (but not MONITOR.FIT itself which does not seem to contain HR data) and using the FitToCSV tool in the FIT SDK to "read" it.
The HR data points have some sort of relative timestamp in seconds (might be seconds since file creation?) associated with each value so easy to difference consecutive points to work out the sampling rates if not so easy the actual time of day.

I have Every Second recording on but not Smart but can't see that is relevant here as that setting is only used, AFAIK, when recording activities. I was told by Garmin that 24x7 HR recording is "Smart" and clearly looking at the "random" sampling intervals that seems to be true. Correlating with my sleep movement, it seemed to take more points when I was "moving" as opposed to deep sleep which makes a lot of sense.

Had not spotted you can export 24x7 data from GC, will have to look again. How do you do that?

Of course HR broadcasting drains the battery. It consumed about 25% of a full charge overnight and after some more "broadcasting" on the way to a run then a 20 mile run, it said was down to 22% when put on USB. Not sure yet whether was "broadcasting" makes it take more points but at least you can see your HR and time of day whenever you want in this mode...

I have Every Second recording on but not Smart but can't see that is relevant here as that setting is only used, AFAIK, when recording activities. I was told by Garmin that 24x7 HR recording is "Smart" and clearly looking at the "random" sampling intervals that seems to be true. Correlating with my sleep movement, it seemed to take more points when I was "moving" as opposed to deep sleep which makes a lot of sense.

True, but I found it to be more sensitive to movement on the old fw.

Had not spotted you can export 24x7 data from GC, will have to look again. How do you do that?

Just go to the daily overview, where you can look at your heart rate and then from the top "settings" icon choose export original. Will give you a zip with a lot of fit files, so finding the right one can take a while, especially because of the misterious timestamp format.

I recorded a macro in excel that does the csv transformations (text in columns, filtering, adding up timestamps and converting to a readable datetime) on the click of a button.

Anybody else tried with HR broadcasting on yet? I seem to be seeing a lot more points today (with this on a fair bit) compared to previous days.